Lett
to the Editor
6. Compendio Estadistico 1965-1974. Managua, Nicaragua: Banco Central de Nicaragua; 1975. 7. Mimeo report. Managua, Nicaragua: Oficina Ejecutiva de Encuestas y Censos; 1977. 8. Nicaragua: 10 Anos en Cifras. Managua, Nicaragua: Instituto Nacionale de Estadistico y Censos; 1989. 9. Anuaano Estadistico de Nicaragua. Managua, Nicaragua: Government of Nicaragua; 1976. 10. Seminario Sobre Eradicacion de la Malaria. Washington, DC: Pan American Health Organization; 1965. PAHO Publicacion Cientifica #118. 11. Grosse RN. Interrelation between health and population: observations derived from field experiences. Soc Sci Med 1980;14c: 99-120. 12. See Wolfe BL, Behrman I. Detenninants of child mortality, health, and nutrition in a developing country. JDev Econ. 1982;11:
163-193. 13. See CaldwellJC. Routes to decreased mortality in poor countries. Popul Dev Rev. 1986;12(2):171-219. 14. Bell D, Reich M, ed. Health, Nutidion, and Economic Cnses. Dover, Mass: Auburn House; 1986. 15. Chen LC. Coping with economic crisis: policy development in China and India. Health Policy Plann 1987;2(2):138-149.
Sanmford and Colleagues Respond It is understandable that Richard Garfield is skeptical of our suggestion that improved access to health care was the factor most likely to have brought about the sharp fall in child mortality that began in Nicaragua in the mid-1970s. It has proven remarkably difficult to demonstrate such an impact in most parts of the world' (although Costa Rica may be an exception2), and despite massive assistance from the United States Agency for Intemational Development, it is generally felt that one of the reasons for the downfall of Somoza's regime was its lack of investment in social programs. Indeed, these were our own prejudices at the time that we embarked upon this work. In fact, it was only after carefully eliminating alternative hypotheses (including most of those put forward by Garfield) that we finally accepted the concept that improvements in child health, at least in recent decades, are not inevitably the outcome of interventions or developments extemal to the health sector. Of course, Garfield's problem, like ours initially, is to find a more plausible explanation for Nicaragua's breakthrugh in child mortality. He starts by suggesting that it may be a delayed effect of economic growth in the 1960s. The troublewith such explanations is that, unless there is a lag 1292 American Journal of Public Health
period that can be justified a priori, it is not possible to explain any change on this basis. Nicaragua's last period of rapid economic growth ended in 1965.3 While it is true that the impact of income growth on health status may be delayed, it is difficult to see how the boom of the early 1960s was still improving infant mortality 13 or 14 years later. The same is true of transport, communications, electricity, and potable water supplies, whose growth and decline in supply closely matched that of the overall gross domestic product.3 Nor does urbanization account for the phenomenon. As our original article showed, the fall in mortality was apparently as rapid in rural as in urban areas and commenced at approximately the same time. In contrast to the supply of energy, water, transport, and communications, the supply of government social services rose steeply in the 1970s and 1980s. This is in accord with our impression that the resources available to primary health care were increasing. That this spending was not entirely soaked up by expansion of hospital-based care is clear from the simultaneous increase in the number of health centers and decrease in the number of hospital beds per capita. Some of Garfield's data should be viewed with caution. Particularly suspect are the data for the number of medical visits per capita. Not only are the figures different from those that we obtained from original sources in Nicaragua, but they also imply that each doctor in Nicaragua was seeing only about six patients per day. It is difficult to see how the number of visits almost tripled between 1978 and 1980 while the number of doctors in the country fell by almost 10%o. Because all of these data depended upon the aggregation of statistics between different institutions, there was obviously plenty of scope for omission and overlap. In fact, the number of patients seen by doctors is probably not the most valid indicator of health care delivery, because it neglects the work of nurses and auxiliary nurses who were staffing most of the ambulatory care units in the country. Regarding Garfield's assertion that only a quarter of the population had access to health care by the end ofthe 1970s, it should be pointed out that a single estimate gives no indication of whether coverage was improving or not. Using Garfield's own figures, the number of births in health institutions (in our opinion one of the better indicators of health service coverage) grewbyan average of 6.0%l
per year from 1974 to 1978 but "only" by 4.2% per year from 1980 to 1986.4(P224) Given the limitations in the available information, any explanation for Nicaragua's interesting trend in child mortality must rely to some degree on speculation, based hopefully on sound theory. However, our study and Garfield's response to it do illustrate both the potential in and the pitfalls of analyzing and interpreting routinely collected health information. [] Peter Sadbrd, MBChB, MSc, MMedSci George Davey Smt, MB BChir, MA, MSc Edward Coye, MBChB Requests for reprints should be sent to Peter Sandiford, MBChB, MSc, MMedSci, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool I3 5QA England.
References 1. Department of International Economic and Social Affairs. Mortality and Health Policy: Proceedngs of the &pert Grop on Mortality and Health Po4y, Rome, 30May to 3 June 1983. New York, NY: United Nations; 1984. Document ST/ESA/SER.A/91. 2. Roxero-Bixby L. Infant mortality in Costa Rica: explaining the recent decline. Stud Fam Plan 1986;17:57-65. 3. Spoor M. Datos macro-economicos de Nicaragua (1960-1986). Managua, Nicaragua: Departamento de Economia Agricola, Universidad Nacional Aut6noma de Nicaragua; 1987. 4. Garfield R, Williams G. Health and Revolution: T7he Nicaraguan Ktperience. Oxford, England: OXFAM; 1989.
htimidafion of CTR-Funded Scientists Clained The July 1991 Joumal carried a Policy Forum discussion by Warnerl and a paper by Cummings et al.2 berating scientists whose work is supported by the Council for Tobacco Research (CTR) for unspecified ethical failures and claiMingthat theworkof those scientists only serves to reinforce doubts in the public mind about the severity of hazards of smoling. Iike many others, my associates and I accept research grants from government, unions, and industry, including CIR. We reported results, with acknowledgments to special grants from CR when appropriate, in more than a dozen leading public health, statistics, epidemiology, and other joumals well known for the thoroughness of their reviews, including theAme,icanJownalof Public Health. Obviously, we had someffiing to say that the reviewers and editors of these journals found worthy of publishing despite scarce journl space and despite acknowledgment to the source of funds.
September 1992, Vol. 62, No. 9
Letten to the Editor
Any survey like those conducted by Wamer and Cummings et al. into the beliefs of special groups of scientists raises the specter of censorship through intimidation. To suppress scientific work because of its consequences is just another excuse for imposing censorship. Moreover, should one really discourage reviews of past research and the implementation of new smoking-related research that is critically oriented? Critical reviews of incorrect and misleading practices in smoking-and-health research serve to highlight erroneous methods and have an important hygienic effect on the conduct of science. At the same time they do not negate results of properly conducted investigations. But, more important, scientists must be free to pursue whatever appears promising to them. For instance, recent observations have shown that smoking is negatively associated with the relative risk of a number of very prevalent and important diseases and with the severity oftheir symptoms, primarily Alzheimer's, Parkinson's, and preeclampsia. Should such relevant research be suppressed because it might increase the sale of cigarettes? (Do scientific disageements really affect the sale of cigarettes?) It is unfortunate that individuals who are strongly dedicated to advancing a social good often appoint themselves as guardians of public morality. I shall seek to publish a full reply to Warner and Cummings et al. in another journal. In the meantime I would be pleased to send a copy of that reply or reprints of our work to interested readers. (Telephone: (604) 733-1348/(604) 6812701; fax (604) 681-2702). Ol Theodor D. Strlin& PhD Requests for reprints should be sent to Theodor D. Sterling, PhD, Faculty of Applied Sciences, School of Computing Science, Simon Fraser University, Bumaby, British Columbia, Canada V5A 1S6.
References 1. Warner KE. Tobacco industiy scientific advisors: Serving society or selling cigarettes? Am JPublic Healtkh 1991;81:839-842. 2. Cummngs KM, Sciandra R, Gingrass A, Davis R. What scientists funded by the tobacco industry believe about the hazards of cigarette smoking. Am J Public Health.
1991;81:894-896.
entists but rather the tobacco industry's motivation in continuing to support the C'R. Simply stated, we believe the goal of tobacco industry management is not to uncover the truth about smoking and health-it is to seil cigarettes. We asserted in our article that the tobacco industry's purpose in funding the CiR is to manipulate and control the flow to the public of scientific information about smoking and health. A recent ruling from a tobacco liability case in New Jersey (Haines vs Liggett group) supports this assertion.2 The judge, in ruling about the admissibility ofdocuments in the case, concluded that the documents contained "explicit admissions" that the tobacco industry had used the CTR to support its legal defense needs. Even today, the tobacco industry continues to deny the causal link between cigarette smoking and lung cancer. In response to a letter writing campaign by a fifth-grade class in Amherst, New York, an R.J. Reynolds spokesman wrote, "the simple and unfortunate fact is that scientists do not know the cause or causes of the chronic diseases reported to be associated with smoking. More scientific research is needed." The letter goes on to cite the industry's support of the CTR. Our article demonstrates that almost all scientists funded by the CTR believe smoking causes disease and would disagree with the assertion in the R.J. Reynolds letter. The fact that a few scientists, such as Dr. Sterling, hold contrary views does not mean that there is significant controversy about tobacco and disease, as the industry wants people to believe, only that there is not unanimity. Neither public health policy nor personal decisions about health need await universal agreement that a substance is dangerous. We had hoped our article would stimulate debate among scientists about the ethical dilemma of accepting funding from the CIR or similar industry-supported entities in light ofthe industry's possible uses of such participation. Apparently, we have succeeded. 0 K. Mihael Cummngs., PhD, MPH Rulwl Scdandma Ronald M. Davis, MD K. Michael Cumings, PhD, MPH, is with
Cwmings et al. Respond Dr. Sterling is incorrect in characterizing our article' as an attempt to discredit
and intimidate scientists funded by the Council for Tobacco Research (CUlR). We do not question the motives of these sci-
September 1992, Vol. 82, No. 9
Rosweli Park Cancer Institute, Buffalo, NY, Russell Sciandra is with the Tobacco Control Prgram of the New York State Health Depart-
ment, and Ronald M. Davis, MD, is with the Michigan Department of Public Health, Lan-
Sing, Mich. Requests for reprints should be sent to K. Michael Cumming, PhD, MPH, Smoking Con-
trol Program, Roswell Park Cancer Institute, Elm and Carlton Sts, Buffalo, NY 14263-0001.
References Cummings KM, Sciandra R, Gingrass A, Davis R. What scientists funded by the tobacco industry believe about the hazards of cigarette smoking. Am J Public Health. 1991;81:894-896. 2. Lew JB. "Fraud!" Judge releases secret tobacco industry documents. Tobacco on TriaL February 28, 1992:1-5. 1.
War7er Responds Obviously, the articles by myself1 and Cummings et al.2 have struck a raw nerve. Unfortunately, it was the wrong nerve, at least as relates to my article. (I would not presume to speak for Dr. Cummings and his colleagues.) The point of my Policy Forum article was to raise consciousness about the tobacco industry's use of its research-funding program as a public relations device, not to challenge the credibility of the research. Indeed, as noted in my artcle, little of that research actually relates directly to the dangers of smoldng, and that which does typically supports the conventionalwisdom on the subject. When the latter is pointed out to the industry, it is the industry itself that challenges the credibility of the science it has funded, invariably insisting that this work is inadequate, that "more research" is needed. I concur with Dr. Sterling's conclusion that my article relates to "the specter of censorship through intimidation." Yet it is not my poll of the CIR Scientific Advisory Board that achieves this heinous outcome; rather, the documented experience demonstrates the industry's ability to intimidate its scientific consultants. The tobacco industry employs its economic muscle to intimidate a wide variety of individuals and institutions throughout our society, including legislators, the media,3 and, apparently, elements of the scientific establishment as well. If Dr. Sterling will take the time to reread my article, he will find only a single call to action. That is the seemingly modest suggestion that the C1R Scientific Advisory Board (SAB) issue a collective statement similar to that of their Australian counterpart, who wrote that "smoking is an important causative factor in several major diseases."4 All of America's major health and medical organizations have taken this stand publicly. What stops the SAB? If it is tnily independent (and free of intimidation), this would constitute a natural means of dissoiating the science it purportedly represents from the distortionaiy use of the CIR program by the American Journal of Public Health 1293
to the Editor
6. Compendio Estadistico 1965-1974. Managua, Nicaragua: Banco Central de Nicaragua; 1975. 7. Mimeo report. Managua, Nicaragua: Oficina Ejecutiva de Encuestas y Censos; 1977. 8. Nicaragua: 10 Anos en Cifras. Managua, Nicaragua: Instituto Nacionale de Estadistico y Censos; 1989. 9. Anuaano Estadistico de Nicaragua. Managua, Nicaragua: Government of Nicaragua; 1976. 10. Seminario Sobre Eradicacion de la Malaria. Washington, DC: Pan American Health Organization; 1965. PAHO Publicacion Cientifica #118. 11. Grosse RN. Interrelation between health and population: observations derived from field experiences. Soc Sci Med 1980;14c: 99-120. 12. See Wolfe BL, Behrman I. Detenninants of child mortality, health, and nutrition in a developing country. JDev Econ. 1982;11:
163-193. 13. See CaldwellJC. Routes to decreased mortality in poor countries. Popul Dev Rev. 1986;12(2):171-219. 14. Bell D, Reich M, ed. Health, Nutidion, and Economic Cnses. Dover, Mass: Auburn House; 1986. 15. Chen LC. Coping with economic crisis: policy development in China and India. Health Policy Plann 1987;2(2):138-149.
Sanmford and Colleagues Respond It is understandable that Richard Garfield is skeptical of our suggestion that improved access to health care was the factor most likely to have brought about the sharp fall in child mortality that began in Nicaragua in the mid-1970s. It has proven remarkably difficult to demonstrate such an impact in most parts of the world' (although Costa Rica may be an exception2), and despite massive assistance from the United States Agency for Intemational Development, it is generally felt that one of the reasons for the downfall of Somoza's regime was its lack of investment in social programs. Indeed, these were our own prejudices at the time that we embarked upon this work. In fact, it was only after carefully eliminating alternative hypotheses (including most of those put forward by Garfield) that we finally accepted the concept that improvements in child health, at least in recent decades, are not inevitably the outcome of interventions or developments extemal to the health sector. Of course, Garfield's problem, like ours initially, is to find a more plausible explanation for Nicaragua's breakthrugh in child mortality. He starts by suggesting that it may be a delayed effect of economic growth in the 1960s. The troublewith such explanations is that, unless there is a lag 1292 American Journal of Public Health
period that can be justified a priori, it is not possible to explain any change on this basis. Nicaragua's last period of rapid economic growth ended in 1965.3 While it is true that the impact of income growth on health status may be delayed, it is difficult to see how the boom of the early 1960s was still improving infant mortality 13 or 14 years later. The same is true of transport, communications, electricity, and potable water supplies, whose growth and decline in supply closely matched that of the overall gross domestic product.3 Nor does urbanization account for the phenomenon. As our original article showed, the fall in mortality was apparently as rapid in rural as in urban areas and commenced at approximately the same time. In contrast to the supply of energy, water, transport, and communications, the supply of government social services rose steeply in the 1970s and 1980s. This is in accord with our impression that the resources available to primary health care were increasing. That this spending was not entirely soaked up by expansion of hospital-based care is clear from the simultaneous increase in the number of health centers and decrease in the number of hospital beds per capita. Some of Garfield's data should be viewed with caution. Particularly suspect are the data for the number of medical visits per capita. Not only are the figures different from those that we obtained from original sources in Nicaragua, but they also imply that each doctor in Nicaragua was seeing only about six patients per day. It is difficult to see how the number of visits almost tripled between 1978 and 1980 while the number of doctors in the country fell by almost 10%o. Because all of these data depended upon the aggregation of statistics between different institutions, there was obviously plenty of scope for omission and overlap. In fact, the number of patients seen by doctors is probably not the most valid indicator of health care delivery, because it neglects the work of nurses and auxiliary nurses who were staffing most of the ambulatory care units in the country. Regarding Garfield's assertion that only a quarter of the population had access to health care by the end ofthe 1970s, it should be pointed out that a single estimate gives no indication of whether coverage was improving or not. Using Garfield's own figures, the number of births in health institutions (in our opinion one of the better indicators of health service coverage) grewbyan average of 6.0%l
per year from 1974 to 1978 but "only" by 4.2% per year from 1980 to 1986.4(P224) Given the limitations in the available information, any explanation for Nicaragua's interesting trend in child mortality must rely to some degree on speculation, based hopefully on sound theory. However, our study and Garfield's response to it do illustrate both the potential in and the pitfalls of analyzing and interpreting routinely collected health information. [] Peter Sadbrd, MBChB, MSc, MMedSci George Davey Smt, MB BChir, MA, MSc Edward Coye, MBChB Requests for reprints should be sent to Peter Sandiford, MBChB, MSc, MMedSci, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool I3 5QA England.
References 1. Department of International Economic and Social Affairs. Mortality and Health Policy: Proceedngs of the &pert Grop on Mortality and Health Po4y, Rome, 30May to 3 June 1983. New York, NY: United Nations; 1984. Document ST/ESA/SER.A/91. 2. Roxero-Bixby L. Infant mortality in Costa Rica: explaining the recent decline. Stud Fam Plan 1986;17:57-65. 3. Spoor M. Datos macro-economicos de Nicaragua (1960-1986). Managua, Nicaragua: Departamento de Economia Agricola, Universidad Nacional Aut6noma de Nicaragua; 1987. 4. Garfield R, Williams G. Health and Revolution: T7he Nicaraguan Ktperience. Oxford, England: OXFAM; 1989.
htimidafion of CTR-Funded Scientists Clained The July 1991 Joumal carried a Policy Forum discussion by Warnerl and a paper by Cummings et al.2 berating scientists whose work is supported by the Council for Tobacco Research (CTR) for unspecified ethical failures and claiMingthat theworkof those scientists only serves to reinforce doubts in the public mind about the severity of hazards of smoling. Iike many others, my associates and I accept research grants from government, unions, and industry, including CIR. We reported results, with acknowledgments to special grants from CR when appropriate, in more than a dozen leading public health, statistics, epidemiology, and other joumals well known for the thoroughness of their reviews, including theAme,icanJownalof Public Health. Obviously, we had someffiing to say that the reviewers and editors of these journals found worthy of publishing despite scarce journl space and despite acknowledgment to the source of funds.
September 1992, Vol. 62, No. 9
Letten to the Editor
Any survey like those conducted by Wamer and Cummings et al. into the beliefs of special groups of scientists raises the specter of censorship through intimidation. To suppress scientific work because of its consequences is just another excuse for imposing censorship. Moreover, should one really discourage reviews of past research and the implementation of new smoking-related research that is critically oriented? Critical reviews of incorrect and misleading practices in smoking-and-health research serve to highlight erroneous methods and have an important hygienic effect on the conduct of science. At the same time they do not negate results of properly conducted investigations. But, more important, scientists must be free to pursue whatever appears promising to them. For instance, recent observations have shown that smoking is negatively associated with the relative risk of a number of very prevalent and important diseases and with the severity oftheir symptoms, primarily Alzheimer's, Parkinson's, and preeclampsia. Should such relevant research be suppressed because it might increase the sale of cigarettes? (Do scientific disageements really affect the sale of cigarettes?) It is unfortunate that individuals who are strongly dedicated to advancing a social good often appoint themselves as guardians of public morality. I shall seek to publish a full reply to Warner and Cummings et al. in another journal. In the meantime I would be pleased to send a copy of that reply or reprints of our work to interested readers. (Telephone: (604) 733-1348/(604) 6812701; fax (604) 681-2702). Ol Theodor D. Strlin& PhD Requests for reprints should be sent to Theodor D. Sterling, PhD, Faculty of Applied Sciences, School of Computing Science, Simon Fraser University, Bumaby, British Columbia, Canada V5A 1S6.
References 1. Warner KE. Tobacco industiy scientific advisors: Serving society or selling cigarettes? Am JPublic Healtkh 1991;81:839-842. 2. Cummngs KM, Sciandra R, Gingrass A, Davis R. What scientists funded by the tobacco industry believe about the hazards of cigarette smoking. Am J Public Health.
1991;81:894-896.
entists but rather the tobacco industry's motivation in continuing to support the C'R. Simply stated, we believe the goal of tobacco industry management is not to uncover the truth about smoking and health-it is to seil cigarettes. We asserted in our article that the tobacco industry's purpose in funding the CiR is to manipulate and control the flow to the public of scientific information about smoking and health. A recent ruling from a tobacco liability case in New Jersey (Haines vs Liggett group) supports this assertion.2 The judge, in ruling about the admissibility ofdocuments in the case, concluded that the documents contained "explicit admissions" that the tobacco industry had used the CTR to support its legal defense needs. Even today, the tobacco industry continues to deny the causal link between cigarette smoking and lung cancer. In response to a letter writing campaign by a fifth-grade class in Amherst, New York, an R.J. Reynolds spokesman wrote, "the simple and unfortunate fact is that scientists do not know the cause or causes of the chronic diseases reported to be associated with smoking. More scientific research is needed." The letter goes on to cite the industry's support of the CTR. Our article demonstrates that almost all scientists funded by the CTR believe smoking causes disease and would disagree with the assertion in the R.J. Reynolds letter. The fact that a few scientists, such as Dr. Sterling, hold contrary views does not mean that there is significant controversy about tobacco and disease, as the industry wants people to believe, only that there is not unanimity. Neither public health policy nor personal decisions about health need await universal agreement that a substance is dangerous. We had hoped our article would stimulate debate among scientists about the ethical dilemma of accepting funding from the CIR or similar industry-supported entities in light ofthe industry's possible uses of such participation. Apparently, we have succeeded. 0 K. Mihael Cummngs., PhD, MPH Rulwl Scdandma Ronald M. Davis, MD K. Michael Cumings, PhD, MPH, is with
Cwmings et al. Respond Dr. Sterling is incorrect in characterizing our article' as an attempt to discredit
and intimidate scientists funded by the Council for Tobacco Research (CUlR). We do not question the motives of these sci-
September 1992, Vol. 82, No. 9
Rosweli Park Cancer Institute, Buffalo, NY, Russell Sciandra is with the Tobacco Control Prgram of the New York State Health Depart-
ment, and Ronald M. Davis, MD, is with the Michigan Department of Public Health, Lan-
Sing, Mich. Requests for reprints should be sent to K. Michael Cumming, PhD, MPH, Smoking Con-
trol Program, Roswell Park Cancer Institute, Elm and Carlton Sts, Buffalo, NY 14263-0001.
References Cummings KM, Sciandra R, Gingrass A, Davis R. What scientists funded by the tobacco industry believe about the hazards of cigarette smoking. Am J Public Health. 1991;81:894-896. 2. Lew JB. "Fraud!" Judge releases secret tobacco industry documents. Tobacco on TriaL February 28, 1992:1-5. 1.
War7er Responds Obviously, the articles by myself1 and Cummings et al.2 have struck a raw nerve. Unfortunately, it was the wrong nerve, at least as relates to my article. (I would not presume to speak for Dr. Cummings and his colleagues.) The point of my Policy Forum article was to raise consciousness about the tobacco industry's use of its research-funding program as a public relations device, not to challenge the credibility of the research. Indeed, as noted in my artcle, little of that research actually relates directly to the dangers of smoldng, and that which does typically supports the conventionalwisdom on the subject. When the latter is pointed out to the industry, it is the industry itself that challenges the credibility of the science it has funded, invariably insisting that this work is inadequate, that "more research" is needed. I concur with Dr. Sterling's conclusion that my article relates to "the specter of censorship through intimidation." Yet it is not my poll of the CIR Scientific Advisory Board that achieves this heinous outcome; rather, the documented experience demonstrates the industry's ability to intimidate its scientific consultants. The tobacco industry employs its economic muscle to intimidate a wide variety of individuals and institutions throughout our society, including legislators, the media,3 and, apparently, elements of the scientific establishment as well. If Dr. Sterling will take the time to reread my article, he will find only a single call to action. That is the seemingly modest suggestion that the C1R Scientific Advisory Board (SAB) issue a collective statement similar to that of their Australian counterpart, who wrote that "smoking is an important causative factor in several major diseases."4 All of America's major health and medical organizations have taken this stand publicly. What stops the SAB? If it is tnily independent (and free of intimidation), this would constitute a natural means of dissoiating the science it purportedly represents from the distortionaiy use of the CIR program by the American Journal of Public Health 1293
